January 25, 2013
Allow us to paint a vision of the future for you: After interactions with your favorite companies, no one asks you how you liked those interactions. Your email inbox contains no requests for a few minutes of your time. No one asks you to wait on the phone line to answer a few questions. The word "survey" has vanished from your vocabulary.
Yes, you're living in a survey-free world (cue: heavens parting and angels singing). This is a vision shared by the comedian Bill Maher who made up a "New Rule" on his TV show: "Get Rid Of Customer Satisfaction Surveys." We also heard this from many of the people we interviewed for our report "Top Ways To Combat Survey Fatigue." One interviewee eloquently predicted, "Surveys are like telemarketers: They'll go away." Others are publicly saying, "Goodbye, Surveys."
But will surveys really disappear? Sorry, folks, that's wishful thinking.
The reality is that your company's customer experience measurement and voice of the customer programs rely on surveys for good reason. At the heart of those programs is the customer experience. Forrester defines customer experience as the perceptions your customers have of their interactions with your company. Perceptions are, therefore, the only true measure of customer experience quality. Until there's a (safe) way to look into your customers' brains, you'll need surveys to figure out their perceptions.
But don't get depressed just yet. Surveys may never disappear completely, but we'll certainly rely on fewer of them.
It starts with making current surveys better. Why? With better surveys, you'll increase your response rates. And that means you'll have to solicit customers less frequently to fill out a survey. And you'll make surveys a better experience for your customers. Imagine that! In our new report on survey fatigue, we outline ways companies can entice customers to start surveys in the first place (hint: prove it's worth their time) — and just as important, how they can get them to finish surveys once they've started. For example, make surveys short, with personalized, relevant, and clear questions. Sure, it'll be complicated to bring in the data you already know about a customer, and you'll have to take out some of those pesky questions no one internally actually uses. But your response rates might reach 40% or 50%.
We can also reduce the number of surveys by getting smarter about using existing data sources. Even if we can't use ESP to figure out how customers perceived their interactions with a company, we can get a lot better at making educated guesses. A good guess is all about matching up behavior patterns or unstructured comments with known measures of perceptions.
NewBrandAnalytics, for example, has figured out a way to predict customer ratings in review and social media sites. It looks historically at which star ratings are associated with what sentiment and language in a customer review. Based on that data, it's building models that predict star ratings based on review content and tone, eliminating the need to ask for the rating directly. This kind of approach could reduce how much survey data you need — if social media provides enough and the right kind of feedback for your company. For instance, companies that serve consumers or many small and medium-size businesses can take advantage of social media more easily than companies serving large corporate or institutional customers.
In a similar vein, one business-to-business technology company we interviewed has plans to base proactive service recovery efforts on this type of predictive modeling. It starts with identifying customer profiles and the behavior patterns that typically result in poor feedback from those customers — in other words, red flags or known indicators of a bad experience. Rather than surveying the customers to find out how they liked their experience, the company will be able to predict that those customers had a bad experience. Once it knows that, it wants to proactively reach out to them and make things right.
Making educated guesses is probably not always completely accurate. But it can still provide value while reducing the number of times you have to reach out to your customers.
The reality, though, is that predictive analytics can only take you so far. It's time to swallow the bad news: Surveys are here to stay.